C++統計精確時間

QueryPerformanceFrequency用法
 

精確獲取時間：
 
QueryPerformanceFrequency() - 基本介紹
 
類型：Win32API
 
原型：BOOL QueryPerformanceFrequency(LARGE_INTEGER *lpFrequency);
 
作用：返回硬件支持的高精度計數器的頻率。
 
返回值：非零，硬件支持高精度計數器；零，硬件不支持，讀取失敗。
 
QueryPerformanceFrequency() - 技術特點
 
供WIN9X使用的高精度定時器：QueryPerformanceFrequency()和QueryPerformanceCounter()，要求計算機從硬件上支持高精度定時器。需包含windows.h頭文件。
 
函數的原形是：
 
BOOL QueryPerformanceFrequency(LARGE_INTEGER *lpFrequency);
 
BOOL QueryPerformanceCounter (LARGE_INTEGER *lpCount);
 
數據類型LARGEINTEGER既可以是一個作為8字節長的整數，也可以是作為兩個4字節長的整數的聯合結構，其具體用法根據編譯器是否支持64位而定。該類型的定義如下：
 
typeef union _ LARGE_INTEGER
 
{
 
struct
 
{
 
DWORD LowPart;
 
LONG HighPart;
 
};
 
LONGLONG QuadPart;
 
} LARGE_INTEGER;
 
在定時前應該先調用QueryPerformanceFrequency()函數獲得機器內部計時器的時鐘頻率。接著在需要嚴格計時的事件發生前和發生之後分別調用QueryPerformanceCounter()，利用兩次獲得的計數之差和時鐘頻率，就可以計算出事件經歷的精確時間。
 
測試Sleep的精確時間：
 
#include <stdio.h>
 
#include <windows.h>
 
void main()
 
{
 
    LARGE_INTEGER nFreq;
 
    LARGE_INTEGER nBeginTime;
 
    LARGE_INTEGER nEndTime;
 
    double time;
 
 
 
    QueryPerformanceFrequency(&nFreq);
 
    QueryPerformanceCounter(&nBeginTime);
 
 
 
    Sleep(1000);
 
 
 
    QueryPerformanceCounter(&nEndTime);
 
    time=(double)(nEndTime.QuadPart-nBeginTime.QuadPart)/(double)nFreq.QuadPart;
 
 
 
    printf("%f\n",time);
 
    Sleep(1000);
 
system("Pause");
 
}

再給出我寫的一個Fibonacci的例子：

#include <iostream>
#include <cmath>
#include <ctime>
#include <windows.h>
#include<iomanip>
using namespace std;

unsigned long Fib1(unsigned long n)
{
    return (n == 1 || n == 2) ? 1 : Fib1(n - 1) + Fib1(n - 2);
}

unsigned long Fib2(unsigned long n)
{
    if (n == 1 || n == 2)
    {
        return 1;
    }
    unsigned long m1 = 1, m2 = 1;
    for (unsigned long i = 3; i <= n; i++)
    {
        m2 = m1 + m2;
        m1 = m2 - m1;
    }

    return m2;
}

int main()
{
    LARGE_INTEGER nFreq;

    LARGE_INTEGER nBeginTime;

    LARGE_INTEGER nEndTime;

    double time1 = 0, time2 = 0;
    unsigned long result1 = 0;
    unsigned long result2 = 0;
    unsigned long fib_num = 45;
    cout.precision(10);
    cout.setf(cout.showpoint); //設置為始終輸出小數點後的數字，就是說 a = 3，它也輸出 3.00000 這
    QueryPerformanceFrequency(&nFreq);

    QueryPerformanceCounter(&nBeginTime);

    result1 = Fib1(fib_num);

    QueryPerformanceCounter(&nEndTime);

    time1=(double)(nEndTime.QuadPart-nBeginTime.QuadPart)/(double)nFreq.QuadPart;

    cout<<"Fib1: "<< fixed<<setprecision(8)<<setiosflags(ios::showpoint)<<time1<<endl;

    QueryPerformanceFrequency(&nFreq);

    QueryPerformanceCounter(&nBeginTime);

    result2 = Fib2(fib_num);

    QueryPerformanceCounter(&nEndTime);

    time2=(double)(nEndTime.QuadPart-nBeginTime.QuadPart)/(double)nFreq.QuadPart;

    cout<<"Fib2: "<< fixed<<setprecision(20)<<setiosflags(ios::showpoint)<<time2<<endl;

    cout<<"Fib2 cost "<<time2*(unsigned long)100/time1<<"% time of Fib1."<<endl;
    //cout << "Hello world!" << endl;
    return 0;
}

C++ Primer Plus 第6版 中文版 清晰有書簽PDF+源代碼 http://www.linuxidc.com/Linux/2014-05/101227.htm

讀C++ Primer 之構造函數陷阱 http://www.linuxidc.com/Linux/2011-08/40176.htm

讀C++ Primer 之智能指針 http://www.linuxidc.com/Linux/2011-08/40177.htm

讀C++ Primer 之句柄類 http://www.linuxidc.com/Linux/2011-08/40175.htm